Flow AI API

💰 Bảng Giá Tạo Ảnh AI

🚀 Bắt đầu nhanh

Bước 1: Tạo ảnh

curl -X POST /public/api/v1/text-to-image \
  -H "Content-Type: application/json" \
  -H "X-API-Key: " \
  -d '{"prompts": ["A cute cat sleeping"}'

# → {"job_id": "gw-abc123", "status": "queued"}

Bước 2: Kiểm tra trạng thái (polling mỗi 3-5s)

curl -H "X-API-Key: " \
  /public/api/v1/jobs/{job_id}

# → {"status": "completed", "progress": 100}

Bước 3: Download kết quả

# Download ảnh
curl -H "X-API-Key: " -o output.png \
  /public/api/v1/jobs/{job_id}/image

# Download video
curl -H "X-API-Key: " -o output.mp4 \
  /public/api/v1/jobs/{job_id}/video

Bước 4: Xem thư viện

curl -H "X-API-Key: " /public/api/v1/my-library
curl -H "X-API-Key: " /public/api/v1/my-jobs
curl -H "X-API-Key: " /public/api/v1/usage

📡 Tất cả Endpoints

📤 Upload ảnh

Upload ảnh jpg/png/webp (max 20MB). Trả về path dùng cho R2I, I2V, R2V.

🎨 Text → Image (T2I)

🖼️ Reference → Image (R2I)

🎬 Text → Video (T2V)

📸 Image → Video (I2V)

🎭 Multi-Reference → Video (R2V)

🔄 Polling & Download

📚 Thư viện & Usage

📐 Chọn Aspect Ratio

🎬 Video

🖥️ Hiển thị Ảnh/Video trên Website

Cách 1: Tải về server rồi serve (Khuyên dùng ✅)

// Node.js — download & lưu file
const resp = await axios.get(
  `${API}/public/api/v1/jobs/${jobId}/image`,
  { headers: {'X-API-Key': KEY}, responseType: 'arraybuffer' }
);
fs.writeFileSync(`./public/media/${jobId}.png`, resp.data);
// Frontend: <img src="/media/gw-abc123.png">

Cách 2: Proxy real-time qua Express

app.get('/media/:type/:jobId', async (req, res) => {
  const {type, jobId} = req.params;
  const ep = type === 'video' ? 'video' : 'image';
  const r = await axios.get(
    `${API}/public/api/v1/jobs/${jobId}/${ep}`,
    {headers: {'X-API-Key': KEY}, responseType: 'stream'}
  );
  res.set('Content-Type', type==='video'?'video/mp4':'image/png');
  r.data.pipe(res);
});
// Frontend: <img src="/media/image/gw-abc123">

Cách 3: Blob URL (Client-side, không cần backend)

async function showMedia(jobId, el, type='image') {
  const r = await fetch(
    `${API}/public/api/v1/jobs/${jobId}/${type}`,
    {headers: {'X-API-Key': KEY}}
  );
  el.src = URL.createObjectURL(await r.blob());
}
// <img id="img1"> → showMedia("gw-abc123", img1)
// <video id="vid1" controls> → showMedia("gw-abc123", vid1, "video")

💡 Code mẫu hoàn chỉnh

JavaScript

const API = "";
const KEY = "";
const H = {"Content-Type":"application/json","X-API-Key":KEY};

async function waitJob(id) {
  while(true) {
    const j = await(await fetch(`${API}/public/api/v1/jobs/${id}`,
      {headers:{"X-API-Key":KEY}})).json();
    if(j.status==="completed") return j;
    if(j.status==="failed") throw new Error(j.error);
    await new Promise(r=>setTimeout(r,3000));
  }
}

// Tạo ảnh 4K
async function createImage(prompt) {
  const r = await fetch(`${API}/public/api/v1/text-to-image`,
    {method:"POST", headers:H,
     body:JSON.stringify({prompts:[prompt]})});
  const {job_id} = await r.json();
  await waitJob(job_id);
  return `${API}/public/api/v1/jobs/${job_id}/image`;
}

// Ảnh → Video
async function imgToVideo(file, prompt) {
  const fd = new FormData(); fd.append("file", file);
  const up = await(await fetch(`${API}/public/api/v1/upload-image`,
    {method:"POST",body:fd,headers:{"X-API-Key":KEY}})).json();
  const r = await fetch(`${API}/public/api/v1/image-to-video`,
    {method:"POST", headers:H,
     body:JSON.stringify({items:[{image_path:up.path,prompt}]})});
  const {job_id} = await r.json();
  await waitJob(job_id);
  return `${API}/public/api/v1/jobs/${job_id}/video`;
}

Python

import requests, time

API = ""
KEY = ""
H = {"X-API-Key": KEY}

def wait_job(job_id, timeout=600):
    t = time.time()
    while time.time()-t < timeout:
        j = requests.get(f"{API}/public/api/v1/jobs/{job_id}", headers=H).json()
        print(f"  {j['status']} — {j.get('progress',0)}%")
        if j["status"]=="completed": return j
        if j["status"]=="failed": raise Exception(j.get("error"))
        time.sleep(5)

def create_image(prompt):
    j = requests.post(f"{API}/public/api/v1/text-to-image",
        json={"prompts":[prompt]},
        headers={**H,"Content-Type":"application/json"}).json()
    wait_job(j["job_id"])
    img = requests.get(f"{API}/public/api/v1/jobs/{j['job_id']}/image", headers=H)
    open(f"img_{j['job_id'][:8]}.png","wb").write(img.content)

create_image("Một chú mèo cam ngồi bệ cửa sổ")

⚠️ Xử lý lỗi